Virtual storage access method (VSAM) catalogs are used to keep track of the condition and location of VSAM datasets in mainframe computing environments. The catalogs are VSAM key-sequenced data sets (KSDS) and are important because VSAM data sets cannot exist until they are first defined in a catalog. Similarly, VSAM data sets cannot be accessed unless information about the VSAM data sets can first be obtained from the catalog. Among other information, the catalogs store information about the amount of space allocated to different VSAM data sets, the addresses of VSAM data sets, the attributes of VSAM data sets, and statistical information about VSAM data sets.
Because catalogs are themselves VSAM data sets, they are prone to splitting of control areas (CAs) and control intervals (CIs) over time which can lead to I/O performance degradation. This splitting can cause the data in the catalog to become disjointed. This problem is well documented and there are various products that have been developed to alleviate this problem. Each of these products, however, typically requires some period of quiesced activity to the catalog while it is reorganized. This generally inhibits both reads and writes to the catalog during the quiesced period. This means that a client may need to cease work that requires access to the catalog during the quiesced period. This is undesirable in today's around-the-clock processing climate.
In view of the foregoing, what are needed are apparatus and methods to reorganize a catalog in order to improve I/O performance. Specifically, apparatus and methods are needed to remove CA and CI splits and order data sequentially so that the catalog can be accessed in a more efficient manner. Ideally, such apparatus and methods would enable the reorganization to take place in a fast and efficient manner while minimally disrupting I/O to the catalog or VSAM data sets referenced by the catalog.